The lure of silicon

Not all cities can be winners in the race to attract high-tech companies

Site selection factors beyond the control of city planners are important to computer companies, according to a study of 317 cities from 1977 to 1992, published in the February European Economic Review.

PROVIDENCE, R.I. -- Cities nationwide have dedicated considerable public resources to lure high-tech industries. Now researchers at Brown University have found that many site-selection factors which are crucially important to highly mobile computer companies may be beyond the control of city planners.

City size, proximity to other centers of high-tech employment, access to a well-educated work force, and employment in related industries are important factors when computer companies decide where to locate, according to the authors of "Spatial evolution of the computer industry in the USA" in the February European Economic Review. Smaller companies - those likely to be looking for a home - are most reliant on such factors.

"Not every city can be a high-tech winner," said J. Vernon Henderson, professor of economics and urban studies and a study author. "Most cities have no computer production. The industry is concentrated in a few locations and even within those areas there is a lot of mobility in a relatively short period of time."

Henderson and Mark A. Beardsell, a graduate student at the time, analyzed statistics on the nation's 317 metropolitan areas for the years 1977 to 1992. During that period, employment in urban areas doubled then halved as the industry exploded with the introduction of personal computers then shrank as manufacturing moved to other countries.

Four of the eight cities that had the largest computer employment in 1977 fell from that list during the time period covered by the study. Of the four cities that remained in the top eight, only San Jose and Orange County had increases. The others - Boston and Minneapolis - lost jobs.

Using economic formulas, the researchers found four characteristics that were important to companies in deciding where to locate:

• Employment in companies making associated electronic components such as circuit boards and semiconductors. It was advantageous for companies to be able to procure supplies locally.
• A medium to large city. Those cities were large enough to provide a diverse work force and had a better chance of supporting other companies in the industry.
• A high proportion of college-educated people. Companies were attracted to a highly skilled work force.
• Location near established centers of high-tech employment.

San Jose, for example, had the largest computer employment between 1977 and 1992. The farther away cities were from San Jose, the less chance they had of attracting high-tech companies. They also stood a greater chance of losing existing companies because companies in or near San Jose had better access to information and suppliers than geographically isolated companies.

Other measures such as average real annual wages in private sector employment or property taxes had no consistent effect on a city's chances of attracting high-tech industry, researchers said.

The industry has already come and gone in several cities, Henderson said, illustrating the randomness of the location process. Factors beyond a city's influence, such as movement of production to other countries or corporate restructuring, contributed to these "flash-in-the-pan" cities: Colorado Springs, Colo; Oxnard, Calif.; Manchester, N.H.; Rochester, N.Y.; Spokane, Wash.; Washington, D.C.; and Tuscon, Ariz.

"People have a sense that the South and West have been big winners in the high-tech race," said Henderson. "It is clear that San Jose is out there as the big winner, but there is a lot of movement elsewhere that will likely continue."

The study was funded by the National Science Foundation and used information from the U.S. Bureau of the Census. The industry's mobility had not previously been widely researched, said Henderson.